Weather Forecast


Patrick Guilfoile: Are our friends really family?

One of life's mysteries is why we end up with the friends that we do. Certainly chance and circumstance play a role. Shared interests and common values are other attributes that are often held in common with friends. Yet a question arises about whether there might be other factors involved. For example, could genetic similarity play a role in determining friendships?

Researchers from Yale University and the University of California-San Diego decided to explore this question. They used detailed information from the Framingham Heart Study, a decades-old investigation into many aspects of the health of a group of people from Framingham, Mass.

These individuals who participated in the study had data collected both on their pattern of friendships and their detailed genetic markers. Consequently, it was possible to determine whether friends shared more genes than individuals who were not friends.

There were a number of complications in interpreting this information, however. For example, people who are friends might be more likely to share ethnicity so in that way, they would have more genetic similarity than expected by chance. To overcome these confounding factors, the researchers used a series of sophisticated statistical techniques to eliminate potential bias in their data.

For example, the researchers took great pains to verify that the friends they study didn't share relatives. They also took steps to ensure factors such as common ethnicity didn't bias their findings.

The results of their analysis were striking. One key finding was that friends have a genetic relatedness equivalent to fourth cousins. There was also an indication that we may select friends based on specific genetic traits. For example, friends are more likely to share similar genes related to our sense of smell. This suggests that friends have a greater ability to detect the same odors compared to individuals who are strangers.

At the other end of the spectrum, there were other genes that were shared between friends less frequently than expected. These included genes for immune system function. It was suggested that this lack of similarity might be useful in preventing the spread of disease, as a person's friends might be less susceptible to certain diseases. This could make it less likely that a friend would pass along a disease, or come down with an illness that had struck another friend.

Based on this information, the researchers developed a "friendship score," which predicted the probability that individuals would be friends, based on their common genes. They found that the level of predictability was similar to that for schizophrenia or obesity.

Like most research, these finding are preliminary, and subject to revision as new information comes in. However, it is intriguing to think that our friendships might have a genetic connection that includes the smells we detect, and the diseases to which we are susceptible.

More information is available in the article by Nicholas. Christakisa and James H. Fowler "Friendship and natural selection." in Proceedings of the National Academy of Sciences USA available at: